Retrieval effectiveness of enhanced bibliographic records
| DOI | https://doi.org/10.1108/eb047797 |
| Published date | 01 March 1990 |
| Date | 01 March 1990 |
| Pages | 43-46 |
| Author | Martin Dillon,Patrick Wenzel |
| Subject Matter | Information & knowledge management,Library & information science |
RETRIEVAL EFFECTIVENESS OF
ENHANCED BIBLIOGRAPHIC RECORDS
Martin Dillon and Patrick Wenzel
OCLC recently completed a research project to
measure the contribution to retrieval
effectiveness of adding abstracts and tables of
contents to bibliographic records. The most
accepted evaluation measurements in information
retrieval research are recall and precision.
Recall is the percentage of relevant documents in
a collection that are retrieved for some query.
Precision is the percentage of retrieved
documents that are relevant. The Findings
indicate that the addition of content-bearing
information in bibliographic records will
improve the overall retrieval effectiveness of
library catalogs. However, the improvement is
primarily in terms of recall. Precision will
suffer as more content-bearing information is
added to records.
INTRODUCTION
There has been a recent surge in interest in
enhancing bibliographic records with additional content-
bearing information. Richard Van Orden's paper,
"Content-Enriched
Access to
Electronic Information,"
included in this issue of Library Hi
Tech,
surveys the
literature on the topic and summarizes the current
research findings. The research project described in
this paper is an attempt to measure the contribution to
retrieval effectiveness of adding abstracts and tables
of contents to bibliographic records.
Definitions
The most accepted evaluative measurements in
information retrieval research are
recall
and precision.
Recall is the percentage of relevant documents in a
collection that are retrieved for some query. Precision
is the percentage of retrieved documents that are
relevant. Normally, these measurements require that
each document in the collection be judged for relevance
to each query. Because of the size of our test file, a
variation
on
recall and precision
was
used.
Rather than
judge all of the documents for all queries, the docu-
ments were ranked by a measure of association to the
queries, and
the top 50
documents for
each
query were
judged for relevance.
Dillon is
director of
the
Office of Research, OCLC,
Inc.,
Dublin, Ohio; and
Wenzel
is a research associate
in that office.
—
RETRIEVAL EFFFECTIVENESS
OF
ENCHANCED...
ISSUE
31
—
(1990, NO.3) 43
To continue reading
Request your trial